There is a rather long discussion bout pthread rwlock conformance to POSIX 2008, and you can see the details here: https://bugzilla.redhat.com/show_bug.cgi?id=1410052 The summary is that glibc does not comply with POSIX 2008 in that we implement the interfaces but not all the semantics. We need to review the conformance and decide if we will ever implement the requested changes, or at the very least document in the manual that the interfaces are not conforming even though we define _POSIX_PRIORITY_SCHEDULING to a non-zero value. The worry is that defining _POSIX_PRIORITY_SCHEDULING to a zero value, but still defining the functions (as Eric Blake suggests and Florian comments on) will cause no end of trouble with code breaking because the functions are not defined. In summary: - Fix the documentation. - Potentially fix hte interfaces. - Potentially define a new value in POSIX for the glibc/linux case.